Arnold Kling  

Happiness Research Still Makes Me Unhappy

Third Rail Issues... Outsourcing Muddles...

Tyler Cowen acknowledges,

Unlike Arnold Kling, however, I do not reject the implications of happiness research altogether.

The ever-excellent Michael of has now come forward and offered a good summary of what happiness research implies...

"Everyone seems to have a pre-programmed "set point" for happiness -- a level of happiness they're genetically programmed for, and to which they'll always tend to return. There isn't much that can be done to change this set point.


My view is that happiness research implies Nothing. Zero. Zilch. Nada. I believe that you do not learn about economic behavior by watching what people say in response to a survey.

When someone asks me how happy I am, what is the basis of my response? Am I looking at a "stock" of happiness (my cumulative happiness over my lifetime, sort of like wealth), a flow of happiness (how I have felt over the past year), or a snapshot of happiness (how I feel this minute)? Am I answering the question the same way as another respondent? Am I even using the same definition if you ask me at two different points in time?

People are not given clear instructions as to how to answer the question, "How happy are you?" In the absence of a precise definition, suppose that I answer the question by giving the ratio of how I feel this week to the average way that I have felt in the past 6 months. In that case, my happiness will always return to a "set point." If I win the lottery, then 6 months later I feel great, but I have felt great for six months, so the ratio of how I feel to how I felt in the last six months is back to where it was, and my reported happiness goes back to where it was before I won the lottery.

Taking out the trash and mowing the lawn would make you very unhappy if yesterday you played golf with your buddies and they are going out on the links again today. But you would be delighted if you had just spent three weeks laid up with a painful injury and now you finally had the strength to do household chores.

My point is that reported happiness is all about comparisons. To report how happy I am, I have to make that report in comparison to something else--how I felt a month ago, or how I imagine someone else feels, how I imagine I'll feel tomorrow, or something. Because the survey questions do not specify the comparison that the respondent is supposed to make, we have no idea what the answers mean.

You can attempt to read something into the data that is reported by "happiness research." You can also read something into tea leaves or goat entrails. Scientifically, there is not much difference.

For Discussion. Why should I take happiness surveys more seriously than goat entrails?

Comments and Sharing

CATEGORIES: Revealed Preference

TRACKBACKS (1 to date)
TrackBack URL:
The author at Catallarchy in a related article titled Arnold Kling Agonistes writes:
    Arnold Kling doesn't like happiness research, and gives a rousing and rather Austrian explanation why not. My point is that reported happiness is all about comparisons. To report how happy I am, I have to make that report in comparison to something... [Tracked on September 8, 2004 12:40 PM]
COMMENTS (14 to date)
SomeCallMeTim writes:

I'm not sure why this is any different than a medical claim that "people felt physical pain when they did [x]." We do not, however, doubt the validity of most medical claims about pain.

Paul N. writes:

"I believe that you do not learn about economic behavior by watching what people say in response to a survey."

Are you serious? Do you believe that you can learn anything from surveys? I think using happiness research to justify retarded economic policies is stupid, but saying economic behavior cannot be judged from surveys is totally off the deep end. It makes our side look desperate and dishonest.

Lawrance George Lux writes:

I am with you, Arnold. If I have ever been happy, I have forgotten when and how it felt. I witness a World where problems accure as fast as population, and see less and less actual action on them. I grew up on a farm, and know a simpler life is the only root for a happy life, yet We daily make life more complicated. I think We all want to experience Death befoe We get there. lgl

Bruce Cleaver writes:

I'll disagree with Michael of

People such as Norman Vincent Peale, Dale Carnegie, and Martin Seligman make an honest living changing the ratio of optimism to pessimism. Here I am using optimism as a proxy for happiness (a weak point, I'll concede).

Arnold Kling, you really ought to read Martin Seligman's works. Optimistic people recover more quickly from psychological traumas, are less discouraged from setbacks, etc.

Bruce Cleaver writes:

One more item - reading Arnold Kling and Lawrance George Lux gives a whole new meaning to the notion of dismal science. Cheer up, guys!

Brad Hutchings writes:

This happiness stuff is going to evolve into the way that policies are sold. It may be goat entrials, and may reek of marketing, but it seems to be the way the game is played.

Three examples from my cursory web browsing this afternoon... c|net has astory about Earthlink being the most liked ISP. Pure marketing -- let's consider that as a baseline. Via NRO, in a story about German anti-Americanism, a statement that 25% of (formerly) West Germans want the Berlin Wall back, and widespread East German feeling that things were better under Communism!!! Implication being that Schroeder foments these feelings so as to further socialize the country.

Third example... I posted a comment about federal overtime rules to a recent TCS essay of yours. Basic point was that such rules which force a 5 day, 8-4/9-5 workday on so many workers cause problems like traffic congestion, fewer personal days for workers, etc. Someone writes back and says that line of argument makes sense (in wanting less regulation of the work week) but doesn't Bush just want to let employers exploit employees? This makes no sense. When I was growing up and my Dad was a manager for the phone company, he would work some pretty long weeks and still have work to take home on the weekend. We lived a pretty comfortable upper-middle-class existence. Most of my friends' Dads worked just as hard and put in just as many hours. I look at my friends today who have kids my age and live a comfortable lifestyle. They never bring home work on weekends and rarely have to "work late". If anything, the historical trend is that as we get more productive, we are self-regulating to not overwork and happier for it. Or put our long hours into our own things (hobbies, startups, etc.). Our laws ought to be consistent with how we seek to arrange our lives to be happier.

The reason to take this happiness stuff seriously is that it is a powerful means of marketing policy. It's just being used (to mis-paraphrase John Kerry) by the wrong people to market the wrong policies right now.

Shaq writes:

Happiness as a metric of life, has been around since the dawn of time. Why else should one live, but to attain happiness.

Acad Ronin writes:

1) I am basically in agreement with you re your skepticism.

2) Still, even if one could meaningfully measure happiness, one runs into two interesting issues in trying to pursue an increase in Gross National Happiness. First, there is the problem of schadenfreude, pleasure at someone else's misfortune. The second is individual utility functions that take relative rank as an argument. (See Fred Hirsch's book, Social Limits to Growth, or Robert Frank's Luxury Fever). Both of these issues suggest that a great deal of happiness is zero sum.

Thomas Blankenhorn writes:
For Discussion. Why should I take happiness surveys more seriously than goat entrails?
Because, as Robert Frank wrote in "How not to buy happiness", the results of happiness surveys have been calibrated against observable measures of (un-)happiness such as suicide. The survey results have been found to correlate well with observations of revealed preferences when both were available.

If the same can be said about goat entrails, it has escaped my attention.

dsquared writes:

What Thomas said; when people get very unhappy, they develop mental illnesses or kill themselves. Happiness surveys track these observable events pretty well.

SomeCallMeTim writes:

Surely what we're talking about here is the granularity with which we can speak about happiness - what are it's constituent parts, what are the appropriate units for those parts, etc. None of which means that one can't make entirely reasonable, rational, "scientific" claims about large-scale movements in happiness and unhappiness. Which, at this point, is all we really care about.

Happy to admit to being nothing more than a fan of following these topics. Still ...

Suppose GDP is increasing. And suppose that the best possible research indicates that people are getting unhappier at the same time. Shouldn't economists be expected to take note, and be concerned?

The idea that economics, of all the social sciences, should be that value-neutral seems a little bizarre, no?

John Thacker writes:

Michael-- it's easy to imagine a situation where people with chronic illnesses are able to live longer, but not in full health, thanks to modern medicine. And so they are demonstrably worse off than previously in their life, and know that they are still certain to die soon. Thus, their happiness is lower than average, since it's worse than most of their life. Dragging illnesses out and keeping fighting them might make the pain last longer for family members, too.

In the aggregate, then, this would cause average happiness to decrease. But, I don't think one would then argue that the few extra months for the sick would be so terrible. (Some people might well agree that dragging things out makes it worse.) Lower average happiness might occur with greater aggregate happiness.

And that's another reason why I have trouble with happiness surveys. It's too easy to construct situations like that.

Joe O writes:

Happiness research seems valid to me. The idea of a happiness set point really resonates with things that happened in my life.

People tend to have a linear mental model of happiness which acts as a motivator. Thinking that you will be twice as happy if you make twice as much makes you work hard.

Happiness maximization does not automatically require high levels of state redistribution of income. Unemployment is one of the few things that does cause long term unhappiness. Too much redistribution could increase unemployment and thus decrease overall happiness.

Comments for this entry have been closed
Return to top